Relative Validity of a Method Based on a Smartphone App (Electronic 12-Hour Dietary Recall) to Estimate Habitual Dietary Intake in Adults.

BACKGROUND: Accurate dietary assessment is key to understanding nutrition-related outcomes and for estimating the dietary change in nutrition-based interventions. When researching the habitual consumption of selected food groups, it is essential to be aware of factors that could possibly affect reporting accuracy. OBJECTIVE: This study aimed to evaluate the relative validity of the current-day dietary recall, a method based on a smartphone app called electronic 12-hour dietary recall (e-12HR), to categorize individuals according to habitual intake, in the whole sample of adults and in different strata thereof. METHODS: University students and employees over 18 years recorded the consumption of 10 selected groups of food using e-12HR during 28 consecutive days. During this period, they also completed 4 dietary records. Once the period was finished, the subjects then completed a food frequency questionnaire (FFQ) and a usability-rating questionnaire for e-12HR. The food group intakes estimated by the e-12HR app, the dietary records, and the FFQ were categorized into sextiles: less than once a week, once or twice a week, 3-4 times a week, 5-6 times a week, once or twice a day, and 3 or more times a day. The 10 selected groups with e-12HR were compared with 4 dietary records and an FFQ reference method, in the whole sample and in different strata thereof: age (years): <25 and ≥25; gender: females and males; occupation: students and employees; smoking: no and yes; physical activity (minutes/week): ≥150 and <150; and body mass index (kg/m2): <25 and ≥25. The association between the different methods was assessed using Spearman correlation coefficient (SCC). Cross-classification and kappa statistic were used as a measure of agreement between the different methods. RESULTS: In total, 203 participants completed the study (56.7% [115/203] women, and 43.3% [88/203] men). For all food groups and all participants, the mean SCC for e-12HR versus FFQ was 0.67 (≥0.62 for all strata). On average, 50.7% of participants were classified into the same category (≥47.0% for all strata) and 90.2% within the nearest category (≥88.6% for all strata). Mean weighted kappa was 0.49 (≥0.44 for all strata). For e-12HR versus RDs, mean SCC was 0.65 (≥0.57 for all strata). On average, 50.0% of participants were classified into the same category (≥47.0% for all strata) and 88.2% within the nearest category (≥86.1% for all strata). Mean weighted kappa was 0.50 (≥0.44 for all strata). CONCLUSIONS: The results indicate that e-12HR generated categories of dietary intake highly comparable with the 2 reference methods in the whole sample and in different strata thereof. The inclusion of photographs to facilitate estimation of the servings consumed generated correlation/agreement data between e-12HR and the FFQ that were similar to a previous study using an older version of the app, which did not include photographs.

Is there any alternative to traditional food frequency questionnaire for evaluating habitual dietary intake?

INTRODUCTION: Dietary assessment methods are an important instrument for nutrition research. Food frequency questionnaires (FFQs) have been the most frequently used dietary assessment tool in epidemiological and intervention studies. There is a great necessity for new methods of determination of habitual dietary intake that overcome the limitations of these traditional methods. OBJECTIVES: The objectives of the present study were to develop a new method, based on an application for mobile phones called e-EPIDEMIOLOGY, designed to record individual consumption data about a short series of foods/drinks, and to compare data collected using this tool with those obtained from a previously validated short paper FFQ. METHODS: University students over 18 years recorded the consumption of certain foods/drinks using e-EPIDEMIOLOGY during 28 consecutive days and then filled out a short paper FFQ at the end of the study period. To evaluate the agreement between both methods, Spearman's correlation coefficient, cross-classification analysis and a weighted kappa statistic were used. RESULTS: One hundred and nineteen participants completed the study (71.4% female and 28.6% male). The mean Spearman's correlation coefficients for food/drink group intake between the two methods was 0.73. The mean percentage of participants cross-classified into categories of "exact agreement + adjacent" was 91.6%. The average weighted kappa statistic was 0.60. CONCLUSIONS: The results indicate that e-EPIDEMIOLOGY has good agreement with the previously validated FFQ short paper. However, it was noted that further testing of e-EPIDEMIOLOGY is required to establish its wider utility.

Is there any alternative to traditional food frequency questionnaire for evaluating habitual dietary intake? / ¿Hay alguna alternativa al tradicional cuestionario de frecuencia de consumo de alimentos para la evaluación de la ingesta alimentaria habitual?

Introduction: Dietary assessment methods are an important instrument for nutrition research. Food frequency questionnaires (FFQs) have been the most frequently used dietary assessment tool in epidemiological and intervention studies. There is a great necessity for new methods of determination of habitual dietary intake that overcome the limitations of these traditional methods. Objectives: The objectives of the present study were to develop a new method, based on an application for mobile phones called e-EPIDEMIOLOGY, designed to record individual consumption data about a short series of foods/drinks, and to compare data collected using this tool with those obtained from a previously validated short paper FFQ. Methods: University students over 18 years recorded the consumption of certain foods/drinks using e-EPIDEMIOLOGY during 28 consecutive days and then filled out a short paper FFQ at the end of the study period. To evaluate the agreement between both methods, Spearman’s correlation coefficient, cross-classification analysis and a weighted kappa statistic were used. Results: One hundred and nineteen participants completed the study (71.4% female and 28.6% male). The mean Spearman’s correlation coefficients for food/drink group intake between the two methods was 0.73. The mean percentage of participants cross-classified into categories of «exact agreement + adjacent» was 91.6%. The average weighted kappa statistic was 0.60. Conclusions: The results indicate that e-EPIDEMIOLOGY has good agreement with the previously validated FFQ short paper. However, it was noted that further testing of e-EPIDEMIOLOGY is required to establish its wider utility (AU)

Introducción: los métodos de evaluación de la dieta son importantes instrumentos para la investigación nutricional. Los cuestionarios de frecuencia de consumo de alimentos (CFCA) han sido los métodos utilizados con mayor frecuencia en los estudios epidemiológicos y de intervención. Hay una gran necesidad de nuevas herramientas que superen las limitaciones de estos métodos tradicionales. Objetivos: los objetivos del presente estudio fueron desarrollar un nuevo método, basado en una aplicación para teléfonos móviles llamada e-EPIDEMIOLOGY, diseñada para registrar los datos de consumo individual de una serie corta de alimentos/bebidas, y comparar estos datos con los obtenidos a partir de un CFCA corto (en papel) validado previamente. Métodos: estudiantes universitarios mayores de 18 años registraron el consumo de ciertos alimentos/bebidas usando e-EPIDEMIOLOGY durante 28 días consecutivos y cumplimentaron un CFCA corto (en papel) al final del periodo de estudio. Para evaluar el acuerdo entre ambos métodos se utilizaron el coeficiente de correlación de Spearman, el análisis de clasificación cruzada y el kappa ponderado. Resultados: un total de 119 participantes completaron el estudio (71,4% mujeres y 28,6% hombres). El coeficiente de correlación de Spearman medio entre ambos métodos fue 0,73. La media del porcentaje de clasificación cruzada en la categoría de «acuerdo exacto + adyacente» fue 91,6%. La media del kappa ponderado fue 0,60. Conclusiones: los resultados indican que e-EPIDEMIOLOGY tiene buena concordancia con el CFCA corto (en papel) previamente validado. Sin embargo, son necesarias más pruebas con e-EPIDEMIOLOGY para establecer su utilidad de forma más amplia (AU)